Dynamic Handwritten Keyword Spotting Based on the NSHP-HMM
Author:
- 1 September 2007
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Ninth International Conference on Document Analysis and Recognition (ICDAR 2007)
- Vol. 1 (15205363) , 242-246
Abstract
This paper presents a keyword spotting system based on the NSHP-HMM. This model allows to dynamically create global word models from letters models, and do not require any writing segmentation. The second section describes our system and its application to a keyword-based handwritten mail sorting task. Next section shows how to divide processing time by 4, using a fix-point arithmetic and a dynamic model desactivation approach based on the natural length complexity. First results are encouraging, particularly for a document-level analysis.Keywords
This publication has 10 references indexed in Scilit:
- Offline recognition of large vocabulary cursive handwritten text2005
- Fast script word recognition with very large vocabulary2005
- Omnilingual segmentation-free word spotting for ancient manuscripts indexation2005
- Fast Two-Level HMM Decoding Algorithm for Large Vocabulary Handwriting Recognition2004
- Word image retrieval using binary features2003
- Large vocabulary off-line handwriting recognition: A surveyPattern Analysis and Applications, 2003
- Using hierarchical shape models to spot keywords in cursive handwriting data2002
- Handwritten month word recognition on Brazilian bank cheques2002
- Cross-learning in analytic word recognition without segmentationInternational Journal on Document Analysis and Recognition (IJDAR), 2002
- HIGH PERFORMANCE UNCONSTRAINED WORD RECOGNITION SYSTEM COMBINING HMMs AND MARKOV RANDOM FIELDS1997